It is said that a rising tide lifts all boats, and so economic growth is generally associated with higher incomes and improved standards of living for everyone. But it’s becoming increasingly clear that the fruits of economic growth are not benefiting everyone. In recent decades, the rising economic tide has floated a small number of luxury yachts, leaving everyone else beached ashore. And we don’t know how it happened.
Many are aware that the share of total income going to the highest earners has been rising. A popular explanation for this divergence is the “skill-biased technical change” hypothesis. Although technical progress is the key to growth, new technologies do not necessarily affect all workers equally. An innovation that improves the productivity of certain workers will generate relatively higher wages for that group.
The information technology revolution is probably the defining feature of the past generation. If, as seems likely, the effect of IT is to make high-skilled workers more productive, then we should see an increase in the wages of those with the highest levels of education. Since high-skill workers already had higher wages, income inequality increases. This is a plausible story, but there are at least two reasons for thinking that it is at best a partial explanation for the increase in inequality.
Firstly, empirical evidence – summarized in a recent Bank of Canada study – suggests that notwithstanding improvements in IT, the Canadian education premium has stayed constant. More important, the gains in income are far too concentrated to be consistent with the skill-biased technical change hypothesis.
Publicly available data sets typically divide the population into five or 10 groups, so we hear of the gains made by the top quintile or the top decile of the income distribution. But the actual compression of income is even worse than what these narratives tell.
McMaster University’s Mike Veall has examined data from the tax files of those at the top end of the income distribution, and his numbers reveal that the concentration of incomes is much more severe than what can be inferred from the publicly available data: most of those who are in the top 10% or 20% have seen income gains that are similar to those at the median.
The income share of the top 10% increased to 40.8% from 33.6% between 1982 and 2007, but when this group is broken down further, it turns out that the share of those in the 90%-95% range remained constant, and that of the 95%-99% increased by only 0.8%. If you take into account taxes and transfers, the share of those between the 90th and 99th percentiles – those with incomes between $65,000 and $170,000 in 2007 – has been falling.
The big winners are at the very top of the income distribution. Those in the top 0.5% – making more than $250,000 per year – have seen their share of total income increase from 5.3% to 10.4%. Unless IT has only improved the productivity of one worker in 200, this concentration cannot be explained by skill-biased technical change. Unfortunately, we have yet to come up with a better theory.
When you make policy recommendations based on an embryonic understanding of a problem, you run the risk of making it worse. For example, consider the proposal to increase income taxes on high earners. This idea makes sense at a superficial level, but a deeper examination suggests that such a tax may have perverse effects. If high earners have enough bargaining power to demand and receive pay increases to cover their increased tax bills, a high-income levy could have the paradoxical effect of making high-income earners relatively better off. If government revenues increase, and if the after-tax incomes of high earners stays the same, then the people who are bearing the burden of a high-income surtax must be those who are lower down in the income distribution.
A recent experiment in the United Kingdom illustrates this point very nicely. In the wake of the financial crisis and public anger at the excesses of highly paid bankers, Gordon Brown’s government imposed a 50% tax on bonuses paid by banks. If the story ended here, it would have had a happy ending. But it didn’t: banks simply doubled the size of the bonus pool so that the after-tax payout remained the same. The tax ended up being paid by the banks’ shareholders, their lower-paid employees and their customers. The only people who were not made worse off were the ostensible targets of the tax.
The extreme concentration of income to such a small number of high earners is unhealthy in a democracy. But a proper remedy must be based on a plausible theory of how high earners managed to acquire the bargaining power to extract such high salaries and to fend off tax increases. And we don’t have one yet.
Stephen Gordon is a professor of economics at the University of Laval in Quebec City.
